This invention relates to an apparatus for the preparation of a compound or an alloy having one component with a substantially higher vapor pressure than the other, preferably for the preparation of a semiconductor compound with elements selected from group III and V of the periodic table, particularly gallium phosphide, in general and more particularly to an improved apparatus which the tube or ampule used in such an apparatus is self supporting between a pair of heating ovens thereby permitting improved cooling.
Apparatus of this general type is known in which the components or elements to be combined to form the compound are situated in a closed horizontal reaction tube with its ends surrounded by hollow cylindrical heating ovens, one oven being provided on each end. The heating ovens are arranged in the axial direction of the tube, one behind the other at a predetermined spacing. At the portion of the tube situated between the two heating ovens, a separate heating device, generally a high-frequency heating coil is provided. The entire system is disposed within a pressure vessel, also referred to as an autoclave, having an internal pressure which can be varied as a function of the pressure inside the tube.
As is well known, gallium phosphide has found much use recently in the manufacture of light-emitting diodes for the visible range because of its large band gap. Polycrystalline gallium phosphide, which is used as the starting material for the manufacture of suitable single crystals, is advantageously synthesized through the reaction of gallium with phosphorus in a closed system.
Typical of one manner of making polycrystalline gallium phosphide is the direct synthesis method using gallium and phosphorus at a temperature of about 1500.degree. C. and a pressure of 6 to 35 bar as disclosed by Frosch and Berick in Journal of the Electrochemical Society, Vol. 108, page 251, 1961. The components of the semiconductor compound are placed in a quartz tube which is arranged in a furnace and is provided with a high frequency heating device. The high-frequency heating device is inductively coupled to a graphite boat located in the tube and containing one of the components. For the reaction, the boat with the semiconductor component, such as gallium, is moved through the inductively heated zone of elevated temperature within the heating device. After an additional pass, dense polycrystalline gallium phosphite containing, at the end of a synthesized bar, free gallium is obtained. The induction coil for the high-frequency heating device is brought into the pressure vessel radially and is therefore not movable in the axial direction of the system. Thus, to obtain the zone melting of the gallium, the tube must be moved and is supported for this purpose in a separate guide tube. [See also Bulletin "TRG 300 Gallium Phosphite Synthesis Systems", published by Metals Research, England.]
At the high temperatures which are required for the synthesis of the semiconductor compound, the strength of the wall of the tube is considerably reduced. Because of this the tube is disposed in a pressure vessel having a pressure set higher than the pressure in the tube. Since the internal pressure of the tube cannot be measured the setting of the necessary counter-pressure in the autoclave over the wide range of possible operating pressures is difficult. It will be recognized that the internal pressure of the tube is influenced by various factors such as the temperature of the phosphorus source and thus the vapor pressure of the phosphorus building up above it. In addition, a varying phosphorus pressure can build up in the reaction tube because of the fact that the reaction does not take place in a completely uniform manner. Furthermore, the phosphorus vapor pressure attainable at a given temperature of the phosphorus source also depends on the degree of polymerization of the phosphorus which, as is well known, constitutes a polymorphous element. Although operating conditions can be chosen so that the phosphorus present is completely evaporated as taught by German Pat. No. 1,029,803, the method disclosed therein, namely that of operating without a base body, cannot be used for larger charges.
In systems of this nature, the reaction tube is generally made of quartz. The pressure stability of hollow quartz bodies first increases with increasing temperature until at a temperature of about 800.degree. C. it is about one-third higher than at room temperature. However, as the temperature goes above this temperature the stability declines and the maximum temperature at which such quartz ampoules can still be used, does not greatly exceed 900.degree. C. In the region of the reaction temperature of gallium mentioned above, the strength of the wall is thus reduced considerably.
One manner of overcoming these problems is disclosed in U.S. application Ser. No. 455,912 filed Mar. 28, 1974 now U.S. Pat. No. 3,991,163 and assigned to the same assignee as the present invention. The method taught therein synthesizes the desired components at a temperature which is considerably below the melting point of the compound produced. As taught therein this is possible if provision is made for the less volatile component to be heated in a reaction zone 1 to 2 centimeters wide to a temperature 100 to 500.degree. C. below the congruent melting temperature of the component produced. At the same time, the pressure of the highly volatile component is set to a predetermined fraction of the decomposition vapor pressure of the compound being produced. Even with this method, however, a reaction temperature is still required which is not much below 1000.degree. C. A further disadvantage is found in this method in that the pressure stability of the quartz tube further decreases as the duration of the pressure stress is extended. Since it is more economical to work with large sized charges it is generally desireable to attempt to prepare bars of the synthesized compound or alloy which weigh several kilograms or more in a single operation. Charges of such size thus require correspondingly longer operating time. This would require, using the previously disclosed method, that the tube be subjected to high pressure for a longer time. In addition, in carrying out such a process, the danger of contamination of the reaction product at high temperatures exists. In view of these difficulties, the need for an improved apparatus of this type becomes evident.